Hi Jakob, 

I would recommend reading up a bit more about clinical trials, attending a course, or speaking to a statistician who is experienced in clinical trials research. Doing this before your study will be advantageous in the long run!

There are two possible, relatively straightforward randomisation methods that you could implement (stratified blocks, or minimisation, depending on your total number of stratification factors). 

Take a look at the trials on trial series by MJA. 

One important consideration in the analysis of data from an RCT is the intention to treat (ITT) principle. That is, once randomised, always analysed. Excluding data post randomisation is not appropriate - you have just removed all the benefits of the process of randomisation, and potentially introduced bias into your study. 

I believe there is a paper in the series that talks about ITT. 

Hope this helps!

Kristy 

Hey Kristy,

thank you very much for the advice, I'll have a look.

Jakob

Jakob

RCT are somewhat hard research work . I suggest the research is done in collaboration with other researchers. And as Kristy suggested you need adequate reading on RCT and consultation and some training also. In these trials you need careful selection of candidates  to study, then random allocation to groups and finally, the analysis is affected by drop out from the study.